Biologists reveal the proteins that first see dangerous microbes.

About 3 weeks after a 4-year-old boy visited the emergency room for a nagging eye infection--which doctors easily cured--the boy's mother arrived at the same Canadian hospital. She was infected with the same bacterium, though it had done much more than redden her eyes. She showed signs of shock: plummeting blood pressure and a racing heart.

Physicians suspected sepsis, the destructive overreaction of the immune system to an overwhelming bacterial infection of the blood. They gave the woman massive doses of antibiotics, but it was too late. Just hours after arriving at the hospital, she died.

Both mother and son had been infected with Neisseria meningitides. Why had the boy beat the bacterium but his mother had succumbed? No one knows for sure, but the answer could lie in the two having had differences in certain proteins that stud the surfaces of white blood cells. These so-called toll-like receptors, or TLRs, are the key sensors guiding the body's initial reaction to bacteria, viruses, or fungi. Scientists call this first line of defense the innate immune response.

Over the past 4 years, scientists have found that TLRs respond to microbial features such as a bacterium's cell wall, tail, or even its DNA. In doing so, the receptors trigger white blood cells to engulf and kill infectious microbes or to signal other immune cells to rally to the cause.

Toll-like receptors are the "eyes of the innate immune system," according to Bruce Beutler of the Scripps Research Institute in La Jolla, Calif. "These receptors are the main way that the innate immune system sees pathogens. That had been a big mystery for the last 100 years or so," he says.

The recent discovery of the human versions of these immune sensors has ignited a research frenzy. "The number of papers that have been published on toll-like receptors in the last 2 years is astounding. It's hard to keep up," says Steven R. Kleeberger of Johns Hopkins University School of Public Health in Baltimore.

Insights from TLRs could help immunologists fight sepsis or design vaccines that ward off infections in the first place. Someday, profiling a person's TLR genes "could tell who's at risk to get sepsis," suggests Beutler. Moreover, drugs that block TLRs may stop the immune overreaction responsible for sepsis or thwart other disorders of the immune system.

The field is "moving at the speed of light," says Fabio Re of Dana-Farber Cancer Institute in Boston. To scientists' surprise, they have even found that TLRs show up on fat cells (see "Immune receptors offer new view of fat cells"), may play a role in premature births, and may contribute to lung damage from air pollution.

About a century ago, Russian biologist Elie Metchnikoff offered the first glimpse at the innate immune system when he reported that a thorn stuck in a starfish was rapidly surrounded by amoeba-like cells. In people, white blood cells called macrophages, neutrophils, eosinophils, and dendritic cells lead this front line of immunity.

While these soldiers of the innate immune system can sometimes fend off invading microbes on their own, they often simply hold the pathogens in check until the adaptive, or acquired, immune system comes to the rescue. This later phase of the immune response depends upon so-called T and B cells.

Traditionally, immunology researchers have focused on adaptive immunity, a bias that Douglas T. Fearon of the University of Cambridge in England noted several years ago. "Despite its evolutionary success, innate immunity has been treated with condescension by immunologists. It has been considered a stopgap measure, a temporary expedient for host defense, buying time until acquired immunity took over," he remarked. "In short, innate immunity was unsophisticated, unintelligent, indiscreet, and obsolescent."

With the discovery of TLRs in the 1990s, that viewpoint has itself become obsolete, Fearon and other immunologists argue. They contend that the innate system not only calls forth the adaptive system but provides it with chemical cues that tailor the response of T and B cells. "It's absolutely clear that the triggering of the innate immune system is a required first step in the recruitment of cells of adaptive immunity and in the training of those cells to see things," says David M. Underhill of the University of Washington in Seattle.

The original receptor named Toll was a fly protein first found to have a role in development, not immunity. Then, scientists observed that flies with a mutation in the protein's gene were unable to fight off fungi. Further investigation revealed that the fly protein recognizes a separate insect protein produced during a fungal infection and in turn, activates the fly's immune response.

In 1997, a group led by Charles A. Janeway Jr., a Howard Hughes Medical Institute investigator at Yale University, identified the human version of this receptor. Its activation in a human immune cell triggers the synthesis of molecules called cytokines, which contribute to inflammation, his team found. Moreover, the stimulated immune cell produces another protein that activates quiescent T cells. In short, this work suggested that TLRs link the innate and adaptive arms of the human immune system.

Despite their recent discovery, the receptors appear to have a long history, notes Robert Modlin of the University of California, Los Angeles. Related molecules help plants fight off infections, so TLR ancestors were probably present hundreds of millions of years ago, before plants and animals evolved into separate kingdoms, he explains.

Biologists following up on Janeway's work soon discovered that the human genome contains at least 10 genes encoding TLRs. Investigators quickly confirmed that TLRs in mammals help innate immune cells recognize pathogens.

Beutler's group, for example, studied mutant mice that don't respond to lipopolysaccharide, or LPS, a cell wall component of certain bacteria. The mice are more susceptible to infection by LPS-bearing microbes than other mice are. In 1998, the scientists discovered that their LPS-tolerant mice had mutations in the mouse gene that corresponds to TLR4, the human gene originally identified by Janeway and his colleagues. LPS is also known as endotoxin because it's important in sepsis. When the mammalian body reacts to LPS, the immune system goes into overdrive, releasing a flood of cytokines that produce fever, shock, and often death.

"We realized that toll-like receptor 4 was the endotoxin receptor," says Beutler.

Immunologists now conclude that TLR4 and at least one other protein form a complex on immune cells that binds to LPS and goads the cells into action.

Different TLRs seem to recognize distinct microbial features. For example, TLR2 was once thought to recognize LPS also, but recent studies have linked it instead to other molecules made by a variety of bacteria, including the ones that cause tuberculosis. And in the April 26 SCIENCE, Underhill and his colleagues presented evidence that TLR5 detects flagellin, a protein found only in the whip-like tails of many bacteria.

Last year, a research team led by Shizuo Akira of Osaka University in Japan reported that TLR9 helps the human immune system recognize bacterial DNA. Scientists have sought to exploit this unusual capability by creating vaccines made of pure DNA (SN: 12/4/99, p. 385), but such vaccines tend to work better in mice than in people.

Offering an explanation, a German research team reported in the July 31 PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES that the two species have slightly dissimilar TLR9s and recognize different bacterial DNA sequences.

Sorting out what microbial features a white blood cell's TLR detects will take some time. The complexity of the situation has grown with the discovery that different TLRs may work together.

Last December, Underhill and his colleagues demonstrated that TLR2 can pair with TLR6 to recognize different bacterial proteins from those that TLR2 alone detects. If such pairings are common among the 10 or so known human TLRs, combinations of the receptors may be able to distinguish a far greater variety of microbial parts than single receptors do.

"That changes the thinking [about TLRs] a great deal," says Underhill.

A looming controversy for those working on TLRs is whether the proteins respond to products of the mammalian body itself. Several research groups have reported TLR reactions to products of mammalian cells: molecules called heat-shock proteins and specific fragments of the blood-clotting protein fibronectin.

Why would the receptors react to such products? One explanation may lie in the controversial "danger hypothesis" put forth by Polly Matzinger, an immunologist at the National Institutes of Health in Bethesda, Md. She challenges the traditional notion that the immune system distinguishes between self and nonself and responds directly to the latter. She argues, instead, that the immune system reacts to microbes only when danger signals have been released by infected or injured cells.

Damaged cells discharge heat shock proteins and fibronectin, providing a possible connection between TLRs and Matzinger's ideas. "Besides sensing danger from the outside, they may see danger signals from the self," says Modlin.

Other researchers say it's difficult to establish that TLRs recognize mammalian molecules. "I don't think the evidence is good enough yet," says Beutler.

Whatever molecules induce TLRs to action, researchers are striving to understand what happens next in the immune response. In the Feb. 23 SCIENCE, for example, Modlin and his colleagues showed that activating one TLR leads directly to the death of a microbe. They reported that triggering TLR2 on macrophages infected with the tuberculosis bacterium prompts the immune cells to destroy the pathogens.

Of perhaps greater interest is the link that TLRs provide between the innate and adaptive immune systems. In the simplest scenario, the innate immune cells could send up a generalized distress call for T and B cells. There are hints of greater sophistication, however.

In a paper recently published online by the JOURNAL OF BIOLOGICAL CHEMISTRY, Re and his colleague Jack L. Strominger report that dendritic cells of the innate immune system behave differently depending on whether their TLR2s or TLR4s are triggered. Activate TLR2 on the white blood cell's surface and it releases one broth of immune molecules; stimulate TLR4 and the dendritic cell secretes a different soup.

Re calls the dendritic cell the "orchestral conductor" of adaptive immunity, noting that the chemicals secreted by it shape the manner in which T and B cells act. Consequently, dendritic cells, through the activation of their various TLRs, may inform the adaptive immune system what type of pathogen has invaded and guide the body's overall response.

Figuring out the roles of all the TLRs and learning how to stimulate them artificially may open up new ways to combat infections and improve vaccines. "We will be able to tailor a specific immune response," suggests Re.

On the other hand, in diseases such as sepsis, in which the immune response is the problem, activating TLRs is the last thing physicians would want. Indeed, a promising sepsis drug now being tested in people turns out to block TLR4 from reacting to the LPS in a bacterium's cell wall. "There may be some situations where we want to activate tolls and others where we want to block tolls," says Modlin.

Another area of growing interest is explorations of how people's TLR genes vary. Changes in those genes may help explain why certain people or populations are more or less vulnerable to specific microbes. According to preliminary evidence from Beutler's group, people with mutations in one TLR are more susceptible to the bacterial infections that can cause sepsis than other people are.

"Collectively, the toll-like receptors have a strong influence on the course of infections. They may be the major determinants of susceptibility to many infectious diseases," says Beutler.

Lately, scientists other than immunologists have started paying attention to toll-like receptors. Kleeberger's work focuses on how ozone and other forms of air pollution damage lungs and why some people are more susceptible than others. His team unexpectedly detected a possible role for TLR4 in that damage. When exposed to ozone, mice lacking TLR4 suffer less lung damage than normal mice do, the group reported in the February AMERICAN JOURNAL OF PHYSIOLOGY.

Kleeberger speculates that ozone initially injures cells in the lungs, inducing the release of molecules that activate TLR4 and create more damage through an immune reaction. His group now plans to study whether variations in the TLR4 gene govern a person's susceptibility to ozone.

The course of pregnancy may also be regulated by TLRs, says Jerome F. Strauss III of the University of Pennsylvania Medical Center in Philadelphia whose group reported last March that certain fragments of the clotting protein fibronectin activate TLR4. That's provocative because data by other teams connect increased concentrations of the fragments in a pregnant woman's blood to a higher risk of premature birth. Also, some infections have been associated with preterm births, adding to suspicion that TLR activation influences the timing of birth. "We're looking for better markers to identify women at risk of preterm birth," says Strauss.

Still, it's the immunologists who are most thrilled with their discoveries about TLRs. According to Evelyn Kurt-Jones of the University of Massachusetts in Amherst, who studies how these receptors may recognize viruses, the recent research "has provided some exciting new ways to look at the innate immune system. It has refocused interest on the earliest events in the immune response."

RELATED: Immune receptors offer new view of fat cells

Scientists are rethinking the fat cell, and the newly discovered toll-like receptors may play an integral part in that reevaluation.

Until recently, biologists considered fat cells relatively uninteresting. "For decades, they were pretty much viewed as fat blobs, as inert storage compartments for trigylcerides," says Philipp Scherer of the Albert Einstein College of Medicine in New York.

That worldview shifted in 1994, when scientists discovered that fat cells secrete leptin, a hormone that travels to the brain and regulates feeding (SN: 12/3/94, p. 372). All of a sudden, fat was seen as an endocrine organ.

Now, says Scherer, there's growing evidence that fat cells can do many of the things that immune cells so. For example, they can release and respond to the chemicals called cytokines, which regulate the immune system, particularly the inflammatory reaction.

Last year, Scherer's group reported that fat cells make TLR4. Moreover, when the researchers exposed fat cells to lipopolysaccharide (LPS), one of the bacterial components that activates TLR4, the cells began to make a related receptor, TLR2, and to synthesize secreted immune molecules such as interleukin-6.

Scherer plans to test whether LPS-activated fat cells produce small, pathogen-killing proteins called antimicrobial peptides. He's also curious whether fat cells can swallow and kill microbes as the immune cells called macrophages do. "We basically view the fat cell as an extremely fat macrophage," he says.--J. T.